Use of expected likelihood in sequential model discrimination in multiresponse systems

✍ Scribed by K.B.S. Prasad; M.Someswara Rao

• Publisher: Elsevier Science
• Year: 1977
• Tongue: English
• Weight: 811 KB
• Volume: 32
• Category: Article
• ISSN: 0009-2509
• DOI: 10.1016/0009-2509(77)80237-6

✦ Synopsis

Use of Bayes' theorem for sequenhal model dlscrunmation reqmres the knowledge of the hkehhood of each model However, the lrkehhood, bang a fun&on of parameters, which are considered to be random vanables, IS m Itself a random vanable The standard procedure m such a case LS to use the expected hkehhood m computmg the postenor model probabllltles In tlus work an expression for the expected llkellhood for a multuesponse system IS developed Computational results are presented using multlresponse data for the lirst time to demonstrate the uttity of the expected hkehhood in efficient model dlscnmmatlon m a catalyst fouling system Two discnmmatory design cntena, VIZ, the Box-H111 crttenon and the Roth's cntenon, are compared